Biochemical Pkarmacalogy, Vol. 32. NO. 18, pp. 2671-2677. 1983

Printed in Great Britain. 0006-2952/83 $3.00 + 0.00

@ 1983 Pergamon Press Ltd.

EFFECTS OF THIAZINAMIUM CHLORIDE, PROMETHAZINE AND CHLORPROMAZINE ON

THROMBOXANE Bz SYNTHESIS, PHAGOCYTOSIS AND RESPIRATORY BURST BY RAT ALVEOLAR

MACROPHAGES

JOSEPH CHANG,* SUSAN PIPERATA, MAUREEN SKOWRONEK and ALAN J. LEWIS

Wyeth Laboratories Inc., Department of Experimental Therapeutics, Philadelphia, PA 19101, U.S.A.

(Received 5 November 1982; accepted 16 February 1983)

Abstract-The effects of three phenothiazines, promethazine, thiazinamium chloride and chlorprom- azine, on macrophage function were investigated in rat alveolar macrophages. The study focused on thromboxane B2 (TxB2) synthesis, zymosan phagocytosis, and hexosemonophosphate (HMP) shunt activity in these phagocytes. TxB2 synthesis by resting macrophages was inhibited by thiazinamium chloride and promethazine in a dose-dependent manner. However, chlorpromazine was inhibitory only at 10m3 M. Promethazine treatment of zymosan-activated macrophages led to a concomitant reduction in both phagocytosis and TxB2 synthesis. Thiazinamium chloride inhibited TxBz synthesis but had no effect on the ingestion of zymosan particles. In contrast, chlorpromazine inhibited phagocytosis but not TxBz synthesis except at lo-’ M. The effects of these agents on the formation of TxB2 synthesis from exogenous arachidonic acid were also investigated. Under these conditions where indomethacin, a known cyclooxygenase inhibitor, was inhibitory, promethazine but not thiazinamium chloride inhibited TxB2 synthesis from exogenous arachidonic acid. Treatment of macrophages with promethazine and chlorpromazine but not thiazinamium chloride results in a reduction in the oxidative burst during phagocytosis. The results suggest that the phenothiazines used in this study differ from one another in their actions on macrophage function. Furthermore, the ability of thiazinamium chloride to selectively inhibit arachidonic acid metabolism may contribute to its bronchodilator/antiallergic activity.

Various attempts using both cell-free preparations and intact cells have been made to elucidate the mechanisms of action of phenothiazines on arachi- donic acid metabolism. In spite of this, it is unclear whether these agents have a negative or positive influence on the synthesis of prostaglandins. For example, chlorpromazine has been shown either to inhibit [l-4] or to stimulate [.5] prostaglandin (PG) synthesis by cell-free systems such as bovine seminal vesicles and guinea pig lung homogenate prep- arations. Similarly, equivocal data have been reported using intact cells. In RBL-1 cells and several other established cell lines, Rigas et al. [6] reported that phenothiazines such as chlorpromazine stimu- late PGEz and PGF2,synthesis at doses ranging from 10 to 40 PM, whereas in platelets it was shown that chlorpromazine at 200 PM inhibits arachidonic acid release and reduces the levels of arachidonic acid metabolites, notably thromboxane A: (TxA$ [7,8].

Prostaglandins are known to have a role in the regulation of airway reactivity in the lung [9, 101 and of cellular immune and allergic responses [ll, 121. Alterations in the synthesis of these metabolites could, therefore, potentially affect these regulatory mechanisms in the lung. Furthermore, alveolar macrophages play a critical role in host defense

Address all correspondence to: Dr. Joseph Chang, Department of Experimental Therapeutics, Wyeth Lab- oratories Inc., P.O. Box 8299, Philadelphia, PA 19101.

mechanisms and are known to synthesize and release PGEz and TxAz [13]. Several studies suggest that phenothiazines may act at the level of phagocytic cells. They inhibit phagocytosis [14], the respiratory burst [15, 161, phospholipase C activity [17], super- oxide generation [18, 191, lysosomal alterations [20], and chemotaxis [21].

Recently, thiazinamium chloride, a quaternary analog of phenothiazine, has been shown to differ markedly in its in vivo pulmonary pharmacology from promethazine itself [22,24]. Thiazinamium chloride possesses greater anticholinergic and antial- lergic activity than promethazine without affecting motor activity. Further, unlike promethazine, which is highly lipid soluble, thiazinamium chloride, because of its quaternary nature, does not diffuse easily across cell membranes. Consequently, thia- zinamium chloride is under clinical investigation as an aerosol form since it is poorly absorbed orally.

We now wish to report our investigations exam- ining the effects of thiazinamium chloride and several reference drugs, including other phenothiazines, on TxB2 synthesis, the hexosemonophosphate (HMP) shunt activity, and phagocytosis in rat alveolar macrophages.

METHODS

Materials. Female Wistar rats (125-150 g) were purchased from the Harlan Co., Indianapolis, Ind.

2671

2672 .I. CHANG et al.

Lung lavage fluid was prepared according to the following specifications: 0.85% NaCl, 0.1% dex- trose, 0.1% Na&DTA, 20mM HEPES, lOOunits/ ml penicillin, and 100 ,ug/ml streptomycin. Media 199 (M199), newborn calf serum (NCS), Hanks’ Bal- anced Salt Solution (HBSS), and Dulbecco’s phos- phate buffered saline containing calcium and mag- nesium were purchased from Grand Island Biological CO., New York, NY. Prior to use, NCS was heat inactivated and Ml99 was buffered with HEPES and supplemented with 100 units/ml penicillin and 100 pg/ml streptomycin. [l-‘4C]arachidonic acid (AA) (sp. act. 52.7 m~~mmole), [‘HjTxBz (sp. act. 125 Ci/mmole), [ 1 -r4C]glucose and [6-r4C]glucose (sp. act. 55-60 mCi/mmole) were obtained from New England Nuclear Corp., Boston, MA. Thin-layer chromatography (TLC) plates for analysis used for prostaglandin separations were purchased from VWR Scientific, San Francisco, CA.

The following compounds were obtained from the sources within parentheses: zymosan (Sigma Chem- ical Co., St. Louis, MO), hyamine hydroxide (.I. T. Baker, Phillipsburg, NJ), trichloroacetic acid (TCA) (Aldrich Chemicals, Milwaukee, WI), Hydrofluor (National Diagnostics, Somerville, NJ), Diff-quick (Harleco Co. Gibbstown, NJ), and prostaglandins (The Upjohn Co., Kalamazoo, MI). Thiazinamium chloride and promethazine were synthesized at Wyeth Laboratories, Philadelphia, PA. Chlorprom- azine was a gift from Smith Kline & French, Phi- ladelphia, PA. Ipratropium was donated by C. H. Boehringer, Ridgefield, CT. Indomethacin was a gift from Merck, Rahway, NJ.

Macrophage cultures. Rats were anesthetized with an intraperitoneal injections of sodium pentobarbital (43mg/kg body wt) and exsanguinated by cardiac puncture; their lungs were isolated and lavaged with a total of 50ml of lavage solution. Lungs that exhibited infection or gross pathological changes were not used. The lavage fluid containing macrophages was centrifuged at 300g for 10 min, and the cell pellet was resuspended in serum free M199. Macrophage monolayers were established by incubating 1.5 x 10” cells in petri dishes (35 x 10mm) for 2 hr at 37” in an atmosphere of 95% room air and 5% carbon dioxide. After washing with I-IBSS to remove non-adherent cells, the cul- tures were incubated in serum-free Ml99 with or without zymosan (100 ~~rnl) in the absence or pres- ence of drugs. Before experimentation, cultures con- tained 95.0 i 0.6% alveolar macrophages, as deter- mined by non-specific esterase staining, adherence, and the capacity for phagocytosis. Viability of macro- phages assayed at the end of each experiment by trypan blue exclusion was always >90% unless stated otherwise.

TxB2 radioimmunoassay. Macrophage cultures were incubated with or without drugs for 2 hr since prelimina~ work showed that TxBl production at this time was suboptimal (data not shown). Experi- ments performed at this time thus allowed an assess- ment of both inhibitory and enhancing effects of the drugs, Foliowing incubation, the media were then removed for analysis of TxB: content by radio- immunoassay. TxBz was measured in the culture media by radioimmunoassay according to the

method of Flynn [2.5] with slight modifications. Briefly, [“H]TxBz (10,000 dpm) was equilibrated with TxB2 antiserum (50~1) and known or unknown samples at ambient temperature in 12 x 75 mm poly- styrene culture tubes for 2 hr. Following incubation, the antibody bound [‘H]TxBz was adsorbed to 100 ~1 of 10% dextran-charcoal solution, vortexed, and pelleted at 1200g for 10 min. An aliquot of the supernatant fluid (250 ~1) was then transferred into a scintillation vial and counted for radioactivity in 10 ml of Hydrofluor. The cross-reactions of the anti- serum with other prostaglandins were: 0.0005% (6-keto-PGF,~), 0.0006% (PGFI,). 0.0014% (PGFz,), 0.0006% (PGE2), and0.031% (PGD2). The concentrations of unknown samples were derived from a standard curve for TxB2.

~eierminafion of hexose-nzonophos~hate shunt acrivify. The oxidation of jl-‘“Clglucose was deter- mined with modifications, as described by De Cha- telet and Parce [26]. Briefly, alveolar macrophages were isolated as described and suspended in PBS at a concentration of 5 x lo6 macrophages/ml. All assays were performed in a final volume of 3 ml in 25 ml Erlenmeyer flasks fitted with rubber stoppers to which center wells containing pieces of filter paper saturated with 250 r_il of 1 M hyamine hydroxide were attached. The reaction mixture consisted of 0.2 &i D-[l-“C]- or D-[6-‘4C]glucose, 1.3 X 10m6M glucose, 100 pgirnl of zymosan, and appropriate test com- pounds. The reaction was started by the addition of 5 X lo6 macrophages to the reaction mixture and incubated at 37” for 1 hr. Following incubation, the reaction was stopped by adding 1 ml of 10% TCA to the mixture. After 20 min, the center wells were removed and placed in scintillation vials and counted for radioactivity in 10 ml Hydrofluor.

Phagocytosis assay. The extent of phagocytosis was determined by the following method. Macro- phage monolayers were prepared as described pre- viously. Zymosan particles were added to media of cultures at a final concentration of 100 pg’ml, and phagocytosis was allowed to proceed for 1 hr at 37”. After incubation, the monolayers were washed extensively to remove free zymosan particles, air- dried, fixed with methanol, and stained with Diff- quick. The ingestion of particles by 100 or more cells per culture were evaluated by direct microscopic counting and the phagocytic index was defined by the equation:

% Phagocytosing cells

= Cells that have ingested 2 or more particles x loo

Total number of cells counted

Metabolism of exogenous [“C]arachidonic acid by rat alveolar macrophages. Macrophages were pre- pared as described and incubated with 1 PCi of [‘“Clarachidonic acid for 3 hr at 37” in the absence or presence of the drugs. After the incubations the culture media were rapidly removed and extracted by solvent partitioning with diethyl ether at pH 3.5 to 4.0. The ether extracts were evaporated, residues redissolved. and aliquots applied to thin-layer chro- rnatographic plates (0.2 mm thickness, Merck). The chromatograms were developed in the following sol- vent systems: ethyl acetate-formic acid (80: 1)

Thiazinamium effects on macrophage TxBz synthesis 2673

(R,: PGEz = 0.41; TxBz = 0.57; 6-keto-PGF,, = 0.22; PGF:, = 0.23; arachidonic acid = 0.90) and ethyl acetate-iso-octane-acetic acid-water (110:50:20: 100) (Rf: PGEz = 0.45; TxB2 = 0.63; 6-keto-PGF,, = 0.35; PGFz, = 0.23; arachidonic acid = 1.0). After development, the marker com- pounds were identified by brief exposure of the chro- matogram to iodine vapour. The distribution of radioactivity along the thin-layer chromatogram was determined by cutting each channel into appropri- ately sized strips, eluting with methanol (1 ml) in liquid scintillation vials for 5 min, and then adding scintillation fluid (Hydrofluor).

Statistics. All data were analyzed using Student’s t-test.

RESULTS

Effect on resting macrophages. Rat alveolar macrophages, when placed in culture, synthesized and released low but measurable amounts of TxB2 over a period of 2 hr. Using l-day culture conditions, PGEz was also synthesized but to a lesser extent and was not investigated further in this study. This is in contrast to the mouse pulmonary macrophages where both PGEz and TxB2 are synthesized approx- imately to the same level [13]. Rat alveolar macro- phages treated with doses of phenothiazines that were non-cytotoxic did not differ morphologically from control macrophages. When examined under an inverted tissue culture microscope, drug-treated macrophages remained adherent and retained the characteristic morphology of alveolar macrophages: large (1620pm diameter), round, extended, and containing many vacuoles and granules. Figure 1 illustrates the effects of thiazinamium chloride, pro- methazine, chlorpromazine, indomethacin and ipra- tropium on this basal synthesis. The data indicate

10 -

8-

0 -1 -6 -5 -6 -5 -4 -3 -6 -5 -4 -3 -5 -4 -3 -7 -6 -5 TCI PROMET CHLORPROM IPRATROP INDO

LOG CONCENTRATION (Mb

Fig. 1, Effects of various drugs on TxB2 synthesis by resting macrophages. Macrophage monolayers (1.5 X 106 cells/ plate) were incubated at 37” for 2 hr in the absence or presence of drugs at various concentrations. Following incubation, the culture media were removed and analyzed for TxB2. All values are expressed as the mean TxB2 equiv- alents (ng/ml ? S.E.M.) of at least four determinations. Abbreviations; TCl, thiazinamium chloride; Promet, pro- methazine; Chlorprom, chlorpromazine; Ipratrop, ipratro-

pium; and Indo, indomethacin.

that, when alveolar macrophages were treated with thiazinamium chloride or its parent compound, pro- methazine, the synthesis of TxBz was reduced in a dose-related fashion. Thiazinamium chloride at the highest dose (10-5M) inhibited TxB2 synthesis by more than 70%) and promethazine had the capacity to abolish TxBz synthesis almost completely within the dose range of 10-1-10-3 M. The 1c50 values for thiazinamium chloride and promethazine were 2 x 10m7 M and 2 x lo-‘M respectively. Chlorprom- azine, on the other hand, failed to affect TxBz levels except at 10e3M where a loss of cell viability was evident. It should also be noted that, at this dose, promethazine caused a similar loss of cell viability whereas thiazinamium chloride was non-cytotoxic at all doses. As expected, indomethacin, a known cyclooxygenase inhibitor, reduced TxBl synthesis markedly; at 10m5 M, indomethacin reduced TxB2 synthesis greater than 80%. Ipratropium, a potent anticholinergic agent, had virtually no effect.

Effect on zymosan-induced TxB2 synthesis. Because resting macrophages produced relatively low quantities of TxBr, it was difficult to assess accurately the pharmacological actions of the drugs under study, We, therefore, examined the effects of the various drugs on zymosan-treated macrophages where TxB2 synthesis was increased. Macrophages (1.5 X lo6 cells/plate) phagocytosing zymosan par- ticles synthesized 2 to 2.5 times more TxB2 than resting macrophages (13.3 * 1.4 vs 6.5 2 0.8ng). Under these conditions, the addition of thiazinam- ium chloride at various concentrations caused a well-defined dose-related inhibition (rcso = 6.64 X 10m8 M (Fig. 2A). No loss of cell viability was observed even at 10V3 M. Figure 2B summarizes the results obtained with the various drugs in comparison with thiazinamium chloride. Promethazine inhibited the formation of TxBz by approximately 40% at 10-4M and, although there was a slight decrease at doses of 10e5 and 10m6M, the inhibition was not significant. Chlorpromazine had virtually no effect on zymosan-induced TxB2 synthesis except at 10m3M where there was a significant loss of cell viability. At lo-’ M, chlorpromazine and thiazin- amium chloride had a slight but non-significant stimulatory effect. Ipratropium, under these condi- tions, also failed to have any effect. Predictably, indomethacin caused a marked decrease in TxBz synthesis by these activated cells.

Effect on phagocytosis. One possible explanation for the inhibitory action of the phenothiazines on zymosan-activated TxB2 synthesis was a reduction in the ingestion of zymosan particles. The phago- cytosis data obtained in the presence of the various drugs are presented in Table 1. It is evident that the inhibitory actions of thiazinamium chloride and indomethacin on TxB2 synthesis were not due to the decreased ability of the cells to phagocytose zymo- san. In contrast, promethazine inhibited the inges- tion of zymosan particles within the same dose range that inhibited the formation of TxBz by alveolar macrophages. For example, where TxB1 synthesis was reduced by 39% at 10m4 M, a similar reduction in phagocytosis was observed. Interestingly, although chlorpromazine was inactive against TxB2 synthesis at doses where no loss of cell viability

2674 J. CHANG et al.

“r(b) 20

i

(a)

16

0 10-8 lo- 1 10-o 10-y 10-u lo-’

THlAZlNAMlUMCHLORlDE (MI1

(4 w

16 -

TCI PROMET CHlORi'ROti lPiAiR&' ;NiO- t LOGCONCENTRATION (M)

Fig. 2 (A). Effect of thiazinamium chloride (TCl) on TxB2 synthesis by zymosan-activated alveolar macrophages. Macrophage monolayers (1.5 X lo6 cells/plate) were incubated with various concentrations of TCI in serum-free Ml99 for 10 min prior to the addition of zymosan (100 &ml). Following a further 2-hr incubation, the culture media were removed and analysed for TxB2. Values are expressed as mean TxBz equivalents + S.E.M. of at least four separate macrophage cultures. (B) Effects of various drugs on TxB2 synthesis by zymosan-activated alveolar macrophages. Macrophage monolayers were treated with the various drugs at the indicated concentrations for 10 min before the addition of the stimulus, zymosan (100 .&ml). Follo~ng zymosan addition, the macrophage cultures were further incubated for 2 hr at 37”. After incubation, the culture media were removed and analyzed for TxB2. Values are expressed as TxB2 equivalents k S.E.M. of at least four separate cultures. Abbreviations: TCl, thia- zinamium chloride; Promet, promethazine; Chlorprom, chlorpromazine; Ipratrop, ipratropium; and

Indo, indomethacin.

Table 1. Phagocytosis of zymosan particles by rat alveolar macrophages treated with various drugs*

Drug % Inhibition W) Phagocytic index of phagocytosis

Thiazinamium Cl 10-S iO-” 10-J

Promethazine 10-Z lo-” 10-j

Chlorpromazine 10-x lo-” 10-l

Indomethacin 10-s

Ipratropium 10-a

90 12

92 2 2 0 83 t 4 0 82’10 0

86 * 2 57 2 St 13 2 7-:

66 t 2$ 15 + 8.k 5 2 2+

94 4 1

92 i 1

5 37 86

27 84 95

0

0

* Macrophage monolayers were established in petri dishes and treated with drugs 5 min before the addition of zymosan particles (100 ,&ml). Phagocytosis was allowed to proceed for 2 hr at 37”. The phagocytic index was cal- culated as described in Methods. Results are expressed as the mean phagocytic index 2 S.E.M. of at least four sep- arate experiments. Statistical significance was determined usingstudent’s r-test, by comparing to the zymosan control.

tP < 0.001. $P < 0.05.

occurred, zymosan phagocytosis was inhibited in a dose-related fashion. Indeed, at IO-” M, the phago- cytic process was inhibited by greater than 80%.

Effect on the metabolism of exogenous arachidonic acid. When alveolar macrophages were incubated with [‘4C]arachidonic acid for 3 hr, significant

Table 2. Effects of phenothiazines and indomethacin on exogenous arachidonic acid conversion to thromboxane

B2*

Treatment N CT? [‘;CdF”,“z Inhibition

m (%I

None 5 4194 f 459 Thiazinamium Cl 6 10-x 4135 + 296 0 Promethazine 6 10-d 2486 -+ 403t 41 Chlorpromazine 6 10-’ 4148 t 703 0 Indomethacin 6 10-5 854 i 148t 80

*Macrophage monolayers were established at a cell den- sity of 1.5 x lo6 cells/plate. [“‘C]Arachidonic acid (0.5 PCi) was added to each culture in the presence or absence of drugs and incubated for 3 hr at 37”. Following incubation, the culture media were removed, extracted, and analyzed for radioactive TXBZ as described in Methods. Results are expressed as the mean dpm + S.E.M. N = number of experiments. Percent inhibition was calculated relative to the control TxB2 level. Statistical significance was deter- mined using Student’s f-test.

tP < 0.01. $P < 0.001.

Thiazinamium effects on macrophage TxB2 synthesis 2615

25 = [6-i4C]

ZYMOSAN Ipg/mlt

Fig. 3. Ability of alveolar macrophages to oxidize [l- W]glucose to ‘%Oz via the HMP shunt. Macrophages (5 ~l@cells/flask) were incubated at 37” for 1 hr with [l- 14C I&? lucose or [6-‘4]glucose in the presence or absence of zymosan at various concentrations as described in Methods. After termination of the reaction, the radiolabeled CO2 that had evolved from the reaction was absorbed into hyamine hydroxide and counted for radioactivity. Values are expressed as the mean dpm -t S.E.M. of at least four

separate reactions.

amounts of radiolabeled TxB2 could be found in the medium. Table 2 shows the effects of the various drugs under these conditions. Addition of indo- methacin to the incubation medium markedly reduced the conversion of [14C]arachidonic acid to TxB2. In contrast, thiazinamium chloride at the high- est dose tested (10e3 M) had no effect, whereas pro- methazine inhibited the formation of [‘4C]T~B~ by approximately 40% at 10-4M. Chlorpromazine at the highest dose (1O-4 M), which did not affect cell viability, did not significantly reduce the synthesis of TxBz from exogenous arachidonic acid.

Effect on the hexosemonophosphate (HMP) shunt. We also examined the effect of these drugs on the HMP shunt, a further index of macrophage function

which has been reported to be affected by pheno- thiazines [14, 151. The degree of glucose oxidation via the HMP shunt depends on the ability of macro- phages to oxidize preferentially [l-‘4C]glucose over [6-‘4C]glucose to 14CO~. In contrast, [6-14C]glucose is solely oxidized through the Krebs cycle, whereas [l-‘4C]glucose can be metabolized by either the HMP shunt or the Krebs cycle. It can be seen from Fig. 3 that actively phagocytosing macrophages exhibited an increased capacity to oxidize [l-14C]glucose in response to increasing doses of zymosan. In contrast, there was no increase in Krebs cycle activity through- out the dose range.

Table 3 depicts the effects of the various drugs on the oxidation of [l-‘4C]glucose by rat alveolar macro- phages. The addition of promethazine or chlorprom- azine caused a dose-dependent inhibition of the HMP activity. At lo-“M, both promethazine and chlorpromazine reduced the level of HMP activity to levels even lower than the basal oxidation of glucose by resting macrophages. Thiazinamium chloride and indomethacin, even at a dose where TxB2 synthesis was reduced by greater than 70%, under these conditions were inactive. Ipratropium equally failed to affect the HMP shunt.

DISCUSSION

In the present study we have examined the effects of phenothiazines on a number of macrophage func- tions. The data establish that: (a) both thiazinamium chloride and promethazine, but not chlorpromazine, inhibited TxBz synthesis in resting macrophages in a dose-related fashion; (b) in actively phagocytosing macrophages, thiazinamium chloride also inhibited TxBz synthesis without affecting phagocytosis, whereas promethazine inhibited both processes; (c) metabolism of exogenous [i4C]arachidonic acid in resting macrophages was unaffected by thiazinam- ium chloride and chlorpromazine but inhibited by promethazine; and (d) zymosan-activated oxidative

Table 3. Effects of various drugs on the hexosemonophosphate shunt in rat alveolar macrophages*

Agent Concn

(M) HMP shunt activity (dpm in 14COz formed from [1-“Clglucose)

None Zymosan (Z)

(100 fiLg/ml) Z + TCI Z + promethazine Z + promethazine Z + promethazine Z + chlorpromazine Z + chlorpromazine Z + chlorpromazine Z + indomethacin Z + ipratropium

1.012 + 196

12,861 2 535 10-3 11,212 2.578 lo-’ < loot 10-j 500 t 196+ 10-s 9.119 2 1,186$ 1om2 i lOOf 1o-4 < 100: 10-S 11,955 t 540 10-s 14,792 i- 842 10-3 13,820 + 1,490

*Macrophages were incubated with or without drugs as described in Methods. HMP shunt activity was calculated as the difference between dpm from [ I-I’CJglucose and dpm from [6-‘4C]glucose. Values are expressed as the mean dpm + S.E.M. of at least four separate determinations.

tP < 0.001. $P < 0.01.

2676 J. CHANG ef al.

burst as measured by the HMP shunt was similarly 0-tetradecanoylphorbol-13-acetate (TPA) as a solu- unaffected by thiazinamium chloride but inhibited ble stimulus showed that thiazinamium chloride still by the other phenothiazines. inhibited TxB2 synthesis (unpublished observations).

Our observation that phenothiazines within the dose range lo-’ to lo-’ M caused a reduction in TxBz synthesis is consistent with the inhibitory effect seen in platelets [7,8] but does not resemble the stimu- latory activity described in RBL-1 cells [6]. Certainly, the inhibitory activity of promethazine in actively phagocytosing macrophages can be mainly attributed to a diminution in phagocytosis. The degree of inhibition of both processes in the presence of pro- methazine was similar which suggests that decreased TxBz synthesis may be a consequence of impaired zymosan uptake by macrophages. However, pro- methazine also inhibited the basal synthesis of TxBz synthesis in the absence of zymosan which tends to suggest that inhibition of particle ingestion may not be the sole mechanism of action of the drug. Indeed, in cell-free macrophage homogenates, Wightman et al. [17] have shown that promethazine is inhibitory against phospholipase C (an enzyme involved in free arachidonic acid release), suggesting a direct action on the arachidonic acid metabolic pathway by the drug.

Thiazinamium chloride and promethazine both possess anticholinergic activity in animals [22]. In view of the fact that anticholinergic agents have been reported to inhibit TxB2 synthesis in the guinea pig lung [30], we also examined a specific anticholinergic agent, ipratropium, under our conditions. The data indicate that such inhibitors do not block TxBz syn- thesis by macrophages. The reason for the discrep- ancy is unclear but may suggest that anticholinergics exert their inhibitory actions on TxBz synthesis via a different cell type in the lung.

In direct contrast to promethazine, chlorproma- zine selectively inhibited zymosan phagocytosis by alveolar macrophages in a dose-dependent manner. This inhibitory action resembles the effect of chlor- promazine on phagocytosis of Staphylococcus epi- dermis [27] and zymosan [14] by polymorphonuclear neutrophils. While zymosan phagocytosis was reduced at lower doses, chlorpromazine reduced TxBz synthesis only at 10e3 M in the actively phag- ocytosing macrophages where there was a clear loss of cell viability; even in resting macrophages, the drug was only mildly inhibitory. The reason for its lack of inhibitory activity on TxB2 synthesis is unclear. Studies with chlorpromazine examining prostanoid synthesis have been contradictory. The drug has been reported to be stimulatory in RBL-1 cells but inhibitory in platelets. In cell-free systems such as bovine and ram seminal vesicles, both stimu- latory and inhibitory activities have been reported [l-4]. Such diverse results could perhaps be due to the free radical scavenging nature of chlorpromazine. Depending on existing conditions such as presence of cofactors and substrate conditions, the drug may display opposite effects, as suggested by Gryglewski

P81.

Earlier studies with platelets indicated that pheno- thiazines may interfere with TxB2 synthesis at the phospholipase step prior to the conversion of arach- idonic acid to TxBl by cyclooxygenase. We, there- fore, have investigated the metabolism of exogenous radiolabeled arachidonic acid in macrophages. Under these conditions, indomethacin, a known cyclooxygenase inhibitor, was inhibitory at doses which inhibit the endogenous synthesis of TxB:. In contrast, thiazinamium chloride was ineffective in blocking exogenous TxBz synthesis even at lo-’ M while promethazine was still inhibitory at 10-j M.

These results, therefore, strongly suggest that thia- zinamium chloride inhibits TxBz synthesis by reduc- ing the availability of the endogenous arachidonic acid substrate. It is, however, not known at present whether it is inhibiting phospholipase Al or C since both enzymes have been described in macrophages [17,31] and both are capable of generating free arachidonic acid. We have not ruled out the unlikely possibility that thiazinamium chloride, while inhibit- ing TxB2 synthesis, may enhance the synthesis of other arachidonic acid metabolites and alter the spec- trum of products synthesized by the alveolar macro- phages. It appears that promethazine may also be acting directly on phospholipase although it is also possible that, at higher doses, it affects cyclooxy- genase activity as well.

The third phenothiazine examined in our study is the quaternary promethazine analog, thiazinamium chloride. Owing to its quaternary structure, the drug is not expected to cross cell membranes easily. Such a proposal is supported by the work of Elferink [29], who showed that in open and resealed erythrocyte ghost membranes the quaternary analog of chlor- promazine is restricted to the outside face of the membrane. In spite of this, thiazinamium chloride was more potent than promethazine in inhibiting TxB2 synthesis in both resting and actively phago- cytosing macrophages. Yet, it virtually left the phagocytic process intact. Even at a dose of lo-’ M where there was almost complete cessation of TxBz synthesis, the ingestion of zymosan particles was unaffected. In addition, preliminary data using 12-

Phagocytes such as the alveolar macrophages pos- sess the distinctive feature of undergoing oxidative metabolism via the HMP shunt during phagocytosis to generate oxygen-free radicals for host defense purposes [32]. Since the plasma membrane is a sig- nificant target for phenothiazines, one would expect that the HMP shunt would be affected. Indeed, both promethazine and chlorpromazine reduced HMP shunt activity. This is consistent with the observa- tions of DeChatelet et al. [1.5, 161, where a similar reduction is seen in rabbit alveolar macrophages. In contrast, thiazinamium chloride has no effect on the HMP shunt suggesting that host defense mechanisms may be left intact during treatment with this drug.

The mechanism whereby promethazine and chlor- promazine impair the HMP shunt is unkuown. Con- ceivably, because of the high affinity of these pheno- thiazines for the plasma membrane and the fact that the (NADPH-NADH) oxidase system is also situ- ated at this location, a situation might be created in which the enzyme is either inhibited or made unavail- able for interaction in the HMP shunt. Additionally, a direct consequence of this inhibition would be a reduction of oxygen-free radicals in macrophages

Thiazinamium effects on macrophage TxB2 synthesis 2677

since the reducing equivalents for the formation of these products are generated through the HMP shunt.

In conclusion, the present studies show that the phenothiazines differ both qualitatively and quan- titatively in their biological actions on alveolar macrophage. However, it is possible that phenothi- azines affect macrophage functions non-specifically. This is unlikely since the three macrophage functions examined in this study, i.e. TxBz synthesis, phago- cytosis and HMP shunt, were not affected to the same degree at a particular concentration. There was no direct correlation between the concentration of phenothiazine necessary to demonstrate inhibition of TxB2 synthesis with that needed to affect changes in phagocytosis and HMP shunt activity. Certainly, with regard to thiazinamium chloride inhibitory actions on arachidonic acid metabolism, our working hypothesis that it inhibits early in the pathway would predict an additional inhibitory effect on leukotriene synthesis as well. Our observations raise the inter- esting possibility that the bronchodilator effect of phenothiazines such as thiazinamium chloride may involve modulation of the arachidonic acid cascade, in addition to the antihistaminic, anticholinergic and mast cell stabilizing (antiallergic) properties already reported.

Acknowledgement-We thank Mrs. Maritza Salicrup for typing the manuscript.

REFERENCES

1. R. E. Lee, Prostaglandins 5, 63 (1974). 2. A. Schaefer, M. Kolmos and A. Seregi, Biochem.

Pharmac. 27. 213 (1978). 3. H. Kunze, E. Bohn and G. Bahrke, J. Pharm. Phar-

mat. 27, 880 (1975). 4. P. Krupp and M. Wesp, Experientia 15, 330 (1975). 5. J. Robak, H. Kasperczvk and A. Chvtkowski,

Biochem. Pharmac. i8, 2844 (1979). . 6. V. A. Riaas. H. Van Vunakis and L. Levine. Prost.

Med. 7, l-83 (1981). 7. J. Y. Vanderhoek and M. B. Feinstein, Molec. Phar-

mat. 16, 171 (1979). 8. J. Maclouf, H. de la Baume. S. Levy-Toledano and J.

P. Caen, Biochim. biophys. Acta 711, 377 (1982).

9. A. L. Hyman, E. W. Spannhake and P. J. Kadowitz, Am. Rev. resp. Dis. 117, 111 (1978).

10. J. Svensson, K. Strandberg, T. Tuvemo and M. Ham- berg, Prostaglandins 14, 425 (1977).

11. J. S. Goodwin and D. R. Webb, Clin. Immun. Immu- nopath. 15, 106 (1980).

12. M. E. Goldyne and J. D. Stobo, CRC Crit. Rev. Immun. 2, 189 (1981).

13. C. A. Rouzer, W. A. Scott, A. L. Hamill and Z. A. Cohn, J. exp. Med. 155, 720 (1982).

14. J. G. Elferink. Biochem. Pharmac. 28. 965 (1979). L. R. DeChatelet, P. S. Shirley, P. Wang, L. C. McPhail and J. P. Gusdon, JR., Proc. Sot. exp. Biol. Med. 153, 392 (1976).

1.5.

17.

18.

19.

20.

21.

22.

23.

24.

16. L. R. DeChatelet, D. Qualliotine-Mann, R. Caldwell, C. E. McCall and J. P. Gusdon. Infect. Immunity 7, 403 (1973). P. D. Wightman, M. E. Dahlgren, J. C. Hall, P. Davies and R. J. Bonney, Biochem. J. 197, 523 (1981). H. J. Cohen, M. E. Chovaniec and S. E. Ellis, Blood 56, 23 (1980). D. L. Ochs and P. W. Reed, Biochem. biophys. Res. Commun. 102. 958 (1981). D. Drenckhahan, L: Klel’ne and R. Lullmann-Rauch, Lab. Invest. 3.5, 116 (1976). B. Bjorksten and P. G. Quie. Znfect. Zmmuniry 14, 948 (1976). A. J. Lewis, A. Dervinis and M. E. Rosenthale, Eur. J. Pharmac. 80, 171 (1982). R. P. Carlson, R. J. Capetola, J. M. DiLuigo, A. J. Lewis and D. H. Yu, Fedn. Proc. 40, 1025 (1981). A. J. Lewis, A. Dervinis, R. P. Carlson, M. E. Rosen- thale and W. C. Daniel, J. Allergy clin. Immun. 69, 155 (1982). J. T. Flynn, Adv. Shock Res. 10, 149 (1983). L. R. DeChatelet and J. W. Parce, in Methods for Studying Mononuclear Phagocytes (Eds. D. 0. Adams, P. J. Edelson and H. Koren), p. 477. Academic Press, New York (1981).

25. 26.

27. T. Huutu, Ann. Med. exu. Biol. Fenn. 50. 24 (1972). 28. R. J. Gruglewski,AgentsActions (Suppl. 7); 247‘(1980). 29. J. G. Elferink, Biochem. Pharmac. 24, 2411 (1977). 30. G. Folco, C. Omini, G. Rossoni, T. Vigano and F.

Berti, Eur. J. Pharmac. 78, 159 (1982). 31. W. Hsueh, U. Desai, F. Gonzalez-Crussi, R. Lamb

and A. Chu, Nature, Lond. 290, 710 (1981). 32. F. Rossi, P. Patriaca and D. Romeo, in The Reticu-

loendothelial System, A Comprehensive Treatise (Eds. A. J. Sbarra and R. R. Strauss). p. 153. Plenum Press, New York (1980).